ATOM Award of the Month, October 2018

For this month's ATOM AotM, we visit the medical industry, and examine a technology that seems quite intuitive, but on account of patents and other obstacles, has seen rapid improvement greatly delayed until now. 

Surgery seems as though robotics would be ideally suited for it, since it combines complexity and precision with a great deal of repetition of well-established steps.  The value of smaller incisions, fewer instances of bones being sawed, etc. is indisputable, from qualitative measures such as healing pain, to tangible economic metrics such as hospital stay duration post-surgery.  

Intuitive Surgical released its Da Vinci robot to the market in 2001, but on account of Intuitive's patents, they sustained a monopoly and did not improve the product much over the subsequent 17 years.  Under ATOM principles, this is a highly objectionable practice, even if technically they can still earn a high profit margin without any product redesigns.  As a result, only 4000 such robots are currently in use, mostly in the US.  Intuitive has achieved a market capitalization of over $60 Billion, so it has succeeded as a business, but this may soon change.  Now that Intuitive's patents are finally close to expiry, a number of competitors are ready to introduce ATOM-consistent exponential improvements into the competitive landscape.  

The Economist has a detailed article about the new entrants into this market, and the innovations they have created.  In addition to mere cost-reduction due to smaller electronics, one obvious extension of the robotic surgery model is for each robot to be connected to the cloud, where the record of each surgery trains an Artificial Intelligence to ensure ever-improving automation for several steps of the surgery.  With AI, greater usage makes it improve, and when thousands of surgeries around the world are all recorded, that makes each machine simultaneously better.  As costs lower and unit volume increases, the volume of data generated rises.  As the accumulation of data rises, the valuation of companies capturing this data also rises, as we have seen in most other areas of technology.  

This level of data combined with greater circuitry within the robot itself can also increase the speed of surgery.  When more of it is automated, and the surgeon is doing less of the direct manipulation, then what is to prevent surgeries from being done at twice or thrice the speed?  This enables a much shorter duration of anesthesia, and hence fewer complications from it.  

 

 

Comments

[Geoman]

I had an interesting eye examine recently. They placed me in a darkened room with a machine on my head. The video screen in front of me lit up, and there was my doctor via video chat. She asked me questions while the machine buzzed and whirred at her direction, taking various measurements of my head, looking in my eyes, etc.

At the end she pronounced my eyes healthy, and prescribed me new glasses.

She was sitting in Florida, 2,000 miles away....she told me she saw a new patient every 20 minutes, all over the country, and could do 20-30 patients every day, never leaving her desk. The examine cost next to nothing. The glasses are automatically machined in the back room, and were ready 1 hour after I selected the frames.

That is the future for all medicine I suspect - remote doctors serially seeing dozens of patients all over the country each day using tele presence and medical robots, with solutions crafted and ready to go while you wait. In the case of the doctors, your prescription, new brace, whatever, is at the front counter waiting for you as you check out. Ideally patients would be sorted going in - joint problems? You are hooked to a specialist that sees nothing but that problem all day. the flu? There is a specialist doctor for that, covering every case of flu in the state.

The GP will be a thing of the past.

[Kartik Gada]

Geoman,

Optometric exams might be possible just from a smartphone app tied to the camera before long. Ultra hi-res imaging with built-in magnification matching the database of millions of previous eye exams will be unbeatable. That will be a different ATOM AotM.

Regarding your Florida Optometrist, that is good! But I wonder why international providers haven't completely captured the US market. The cost from India, or even Poland, would be an order of magnitude lower than the US. Basic optometry isn't the same bar of skill as neurosurgery, so US providers should either be completely outcompeted, or one Optometrist like your one in Florida eliminates 20 others just from sheer volume.

With robotic surgery, there is still no network effect and certainly no AI/Cloud layer. The innovation so far is only in precision and small incisions, but much more can be added to that when you have data and all machines connected to the cloud.

[Palamas]

You all are way out of your league with this. Things like Lasik or optometry exams, while really cool and technical, with great emphasis on physics (optics and lens creation) are WORLDS away from the kind of health care anyone expects, or suffers from. They are great innovations in 1 dimensional fields. The human body with its physiology and interaction with the environment for years, including symptoms such as "pain" make futurism in this case funny. All Da Vinci did was help a physician gain a new skill for improved RECOVERY (less invasive procedure). It didn't do anything for replacing the human or cost. Actually, it increased expert skill requirements and thus, cost from both a human and robot point of view. In a world where everyone wants more and has less, this could easily be just as irrelevant as it is relevant.

Yes, I am a physician.

[Kartik Gada]

Palamas,

All Da Vinci did was help a physician gain a new skill for improved RECOVERY (less invasive procedure).

That is huge. Time spent in the hospital and the productivity loss post-surgery has been slashed.

It didn't do anything for replacing the human or cost.

No one said it did. Plus, the article is about the future of robotic surgery. Da Vinci is outdated because it has been protected by patents. It was innovating in 2001, but not now.

[Ed Zimmer]

I'm afraid Palamas may be clinging to the past (an all too common condition). I believe medical care will be undergoing major changes over the next few years - and not just from robotics and enhanced communications. After having benefited from Medicare the past 20 years, I'm a firm believer in Medicare-for-All (and believe its coming soon despite the monied resistance). From discussions I've had through The Economist during the past few years with physicians, I've learned there are two distinct camps - those that absolutely abhor the program and those who have learned to live with it (and actually enjoy the practice, albeit wishing it to be a bit more generous in its fees).

The difference appears to be between those believing that medical care should be about establishing personal relationships with patients, with regular hour-long appointments scheduled two or more times a year (which approach is obviously not economically practical given Medicare's mandated payment rates) - and those who accept those rates and have made medical care a "business", organized for efficiency and making maximum use of lower-skilled employees (nurse practitioners, physician assistants, etc.) and labor-saving equipment. I believe the "business" approach will be the winner. My experience with the "business" physicians has been infinitely more satisfying and assuring than (what I consider as) the old-style practice.

[Palamas]

Ed,

I live in the present and more importantly, IN this system you are talking about. I am in diagnostics. "It will be undergoing"??? It already has my friend. You think there's enough money or time for all of the patients, let alone the boomers, who will crush the system in the US starting in 2025? Regulations on physicians have only gotten greater, education costs only gotten MUCH steeper (with higher interest rates even!) and mid levels are laughable for anyone who actually needs a physician. This is not a knock on them, rather the expert training it takes that the public wants increasingly for free. Also what the hospitals want to make as commodity or worker bee.

The business approach is already the winner.

I recommend you eat right, exercise, and forget physicians all together. What's coming is going to be funny for those who lack discipline and fitness, which is most of America, at this point.

[Palamas]

Kartik,

There is no bright future for robotic surgery. There was a bright surgery for 70 year old retiree age types with prostate cancer ... way back when I was in medical school. Some can still pop wood. The others, well, who gives a shit anyway. They are fat and 70.

Not trying to be crude, but rather funny, because this is the truth.

The emerging market problems with an increasingly strong dollar is the real issue coming upon us in the world ... or the fact that socialism is totally failing and what are the antifa types going to set off in all of these countries ...

[Geoman]

Well, with all due respect Palamas, I disagree.

What is the reason optometric exams are ahead of the curve on this? Same reason cosmetic surgery is ahead. Innovation follows a free market. TFor long stretch of time optomology and cosmetic surgery have largely not been covered by government programs OR health insurance. That has caused them to innovate and compete, lowering prices and improving service radically. They show the potential for innovation to reduce costs.

I don't think optometric exams are necessarily unique - a vast majority of our ailments are routine and do not require the in-presence of a doctor. heck, you could tier the thing up - we forget the FIRST examine is frequently by a nurse, or nurse practitioner. the second is by the doctor. So the nurse screens you, and a second screening is by the tele presence doctor in India. He eliminates all the flu patients, people with routine and easily manageable problems. and orders all the appropriate testing, and refers forward just a very small number of patients to the in-residence doctor. That doctor makes the final call on the serious stuff.

Or we simply step into a machine, and the machine does most of the stuff the nurse does - weight, weight, temperature, BP.....that goes to the tele-presence doctor that asks us about our symptoms....and guided by ever improving sorting algorithms determines what our problem is.

Let's face it - most "doctoring" is an if/then statement guided by statistical analyses of disease symptoms. Do you feel warm? Thirsty?
What is you age? Do your joints hurt? Tele-presence is going to be very useful to doctors, who can guide the process through a decision tree to quickly arrive at the right diagnosis. And the process is self re-enforcing - machine algorithms will track how many of the diagnosis were correct, and suggest additional questions/tests to screen better.

And again, we don't need ALL medicine to go this way - if we can speed 70% of the patients through with a quick and accurate diagnosis, for half the cost, it is going to be a big improvement.

For my own experience, once I got over the strangeness of it all (machine puffing air in my eyes), I thought it was pretty terrific. It was certainly cheap and efficient.

[Ed Zimmer]

Palamas,
Did you look at the Economist article Kartik provided the link to? As one whose whole career has been in computers (from punched card days), I maintain that robotic surgery has a very bright future, just as Kartik summarized. You may be thinking of robots doing the surgery - if so, better to view the robot as being a tool of the surgeon, giving him/her much better precision and consistency (and the ability to deep-learn the surgical details to benefit all surgeons). The robotics development work going on today is mind-blowing (as is the pace of development), in the open-source community as well as the business.

Re quality of medical care, only 10 years ago I had little choice but to consult with a physician and accept what I was told. Today, via the web, I can research medical papers, stay current with clinical trials of interest and connect with others sharing similar experiences. I still consult with physicians when I perceive the need - but I'm a much more knowledgeable (hence, better) patient. And I'm really looking forward to the plethora of new sensors and diagnostic systems that will be coming online over the next decade where the physician I end up consulting will be the world-class expert in the specific ailment.

BTW, I appreciate the physicians' earnings concerns, especially in light of their educational costs. However, I favor not only Medicare-for-All, but free education as advanced as one demonstrates their capability. And don't label that "socialism" - socialism is about ownership of the means of production (of goods and services), neither of which is applicable here. The wealth of a nation resides in the health and education of its people - not in a fiat currency.

[Kartik Gada]

Geoman,

Let's face it - most "doctoring" is an if/then statement guided by statistical analyses of disease symptoms.

Indeed. At any given time, two-thirds of the cars in the hospital parking lot are probably there for something that a rudimentary diagnostic search engine can handle. If a digital photo from the smartphone can interface into the vision algorithm, then most basic GP visits become unnecessary.

The best Doctors will have MORE to do in such a world, not less. Offloading easy diagnostics and prescriptions to AI frees up thousands of Doctors to work on the tougher challenges.

With surgery, smaller incisions are a major and indisputable improvement in their own right.

[Joe Blow]

I find it very hard to believe that robotic surgery doesn't have a bright future. You only have to look at the robots coming out of Boston Dynamics to see that the robotics field is advancing rapidly. Why would surgery be any different? Its more a case of ‘when’ it will happen rather than ‘if’ it will happen. Kartik has nicely explained why progress in this area has been slow and should soon pick up. This is something which I have occasionally wondered about, having read very little about any real innovations in this area in the science news for a long time.

Re smaller incisions - I had my gall bladder removed about 10 years ago and you literally can't see the scars. I had 5 incisions I think, the largest was hidden in the belly button and the rest about a 1/4 inch long. They pump a gas ( C02?) though one of the incisions to enlarge the space inside so they get a better view. This has a downside that some people get what feels like a heart attack a few hours after the op. This happened to a woman opposite me in the post op area and it was interesting and scary to see the emergency team spring into action. 10 people in less than a minute.

Another interesting tidbit is that at least for this type of operation, if you are obese you had to have the op the old way which involved a scar maybe 8 inches long. Something to think about!

[Kartik Gada]

Joe Blow,

if you are obese you had to have the op the old way which involved a scar maybe 8 inches long. Something to think about!

Yet another benefit to not being obese.

I had my gall bladder removed about 10 years ago

The thing is, Intuitive has not improved their robot since then, as they are protected by patents. Someone today would have the same experience (even if it was good by the standards of the time).

Now that the patents are expiring, we will get a wave of innovation that brings us back to the technological trendline. We are currently far below the trendline. Plus, the innovation in surgical robots will create enough volume for the same innovations to flow elsewhere. The cool videos from Boston Dynamics are indicative of this.

[Palamas]

You guys don't know what you don't know in this regard.

"The easy diagnostics" ... where to begin ...

And then the excuse for how they made the small incision and put CO2 in 10 years ago, which is the same as now as you state, but that has to do with patents?

Without humans you get nothing. AI is gonna replace radiology just like radiology is gonna be outsourced to India (said also 20 years ago). LOL

[Joe Blow]

I don’t know if Palamas is trolling or not, but a youtube search for ‘medical diagnostics automation’ has dozens and dozens of videos showing the field rapidly advancing in every way imaginable. Same for ‘ surgical automation’. Still plenty of humans involved though, just a lot more productive. Googling for ‘ai replace radiologist’ shows a pretty vigorous debate with many in the field urging their fellow radiologists not to be too complacent.

Youtube is great for that sort of stuff. I once saw a video on an automated milk factory that produced a few million litres of milk and milk products a week. Barely a human to be seen and most of them were in research and development. In Scotland I visited a distillery that produced several million litres of fine whiskey each year and 2 people ran the whole plant! The other 10 people working there were mainly involved with taking tours.

What has this got to do with surgery? Not much, except that I can’t see how it would be immune from the same forces.

[Geoman]

https://bgr.com/2018/10/15/google-ai-breast-cancer-detection-lyna/

Google’s AI is now better than doctors at spotting breast cancer. Just saying.

And again, we've all been to more than a few doctors, and honestly...let's just say you are much more impressed with each other than the general public is. My experience, consistently, over decades, is the nurse does most of the "data" collection. The doctor breezes in and out. Lucky if you get 10 minutes. Maybe the doctor is doing something important during that brief visit...but I suspect he mostly just looks at the symptoms and plays the odds.

Palamas - I'm sure you are excellent at what you do, but seriously - there are a lot of clinics out there with.. ahem… less than impressive diagnostic abilities. The doctors aren't that good, or are hurried and busy, or distracted by their mountains of debt and hoping to see 20 patients in a day instead of 15.

And what about cost? the doctor shortage? the immense debt and time to become a doctor? how about the grossly inefficient waste of time filling out forms?

I'm not imagining a system that replaces doctors, I'm imagining a system that focuses doctors on solving difficult health problems.

Frankly the system we have today is expensive, archaic and silly and more than a little dysfunctional. As customers we all know it. And I can think of very few things that have improved as little as the routine doctor's visit over the last 30 years. Diagnostics, surgery, vaccines. All have taken leaps forward. But visiting the doctor? call for an appointment. Fill out forms. wait. Nurse takes your vitals. Wait some more. Doctor comes in, distracted, "okay so Mr...Geoman… says hear you have some ringing in you ears, let's take a look. huh..Okay, here is a prescription for some drops, call me in 2 weeks if it doesn't get better.." an hour or two of my time wasted to see a doctor for 10 minutes, tops.

I think your dismissal of the eye exam thing is a bit nearsighted (yes I made the joke). I fail to see how my experience couldn't translate into a general screening visit to a doctor. I walk in the door, sit down in the examining machine, stick my arm in a cuff, and insurance card in the slot, the screen immediately lights up - "I'm doctor X in Florida. What is your problem today? I see you blood pressure is X, your weight is Y, hold still while I examine your ears...hmmmm...your prescription is waiting at the door."

Same result, half the time, half the cost. Doctor treats a 30 people a day without leaving her desk. Less nurse time too.

"Amazon, Berkshire and JPMorgan Chase to team in landmark new health care company." Amazon - isn't that the same company that destroyed Sears and every bookstore in America? But, pah-shaw, they can't possibly improve health care. "The new company’s goal at first will be to target technology solutions to simplify the health-care system." I'm going to go out on a limb and suggest they will implement exactly what I have suggested - a fully data integrated system that relies heavily on integration and AI diagnostics.

[Kartik Gada]

Of course there is a tremendous amount of automation possible in medicine.

The GP visit, the Optometrist, the Dermatologist, and the Radiologist's image-scanning are all obvious examples.

Remember that in the US, both the cost/years to become a doctor, and the pay that doctors earn, are unusually high compared to even other advanced economies.

This still does not mean that doctors vanish. Rather, the better ones will be freed to do higher-tier work for higher pay.

[Geoman]

And the lower end ones get to see a lot more patients. I think with the right equipment and AI the job of a screening doctor could be quite lucrative and even fun.

Nothing will be wasted. I have little doubt that cost and time are huge factors preventing more people from going to the doctor. Just like there is no eye glass shortage, no cosmetic surgery shortage, there will no longer be a health care shortage for anyone. Every town can have at least a doc in the box to do screening.

Instead of making good health care a luxury good, why not make it for the masses?

[Kartik Gada]

Geoman,

Nothing will be wasted. I have little doubt that cost and time are huge factors preventing more people from going to the doctor.

Superb point. I can attest to this myself. High associated costs reduce people's willingness to schedule the entire ordeal. Again, two-thirds of the cars in any big hospital parking lot are probably things that could be automated.

If the AI enabled certain things to be done from home (and of course during off hours), many important visits will no longer be delayed.

[emma]

I am not the pro and understand a little about medical, but this looks revolutionary.

one day will come when disease will be the thing of past

[Geoman]

Make good health no longer a luxury item, but something the masses of people can enjoy. Very few people have gotten poorer by making their services cheaper and more easily obtainable. Most get very, very, rich.

Imagine the best doctors in the country on-call 24/7 to deal with the most challenging cases. How much more satisfying that would be for all concerned? Imagine them helping dozens of people each day.

One of the problems in medicine is the immense liability associated with it. One rule change I'd like to see is a "good Samaritan" principal applied to health care. Basically, any health care dispensed at zero cost has zero liability. This would allow more innovation with charitable heath care, and lower insurance costs. heck, doctors might just elect to do the screening for free, just to avoid the liability. This would allow a lot more innovation into the game, at least at the lower end of the profession.

[Palamas]

Emma, wrong. Anyone who knows about life and biology wouldn't say such a silly thing. The tower of Babel always gets reinvented. We're different though, sure.

Gents, CAD in radiology has been around for more than 30 years. It still is marginal, and that's in the worst diagnostic specialty (mammography, the simplest) in all of radiology. 30 years. Screening mammo should go away for other reasons, but not computer ones, lol.

AI can't hold a CANDLE to a radiologist. It's classic marketing and monetary interest vs. reality. Believe what you want to believe though. Consciousness is flexible and consultative. Computer analysis is 1 dimensional and lacking responsiveness and creativity.

Go ask a surgeon how much he'd like a CT read by an AI. Seriously, you guys literally do not know what you don't know. It's a huge problem here that you can't even fathom, which is why you keep going on and on.

[Kartik Gada]

Palamas,

AI can't hold a CANDLE to a radiologist. It's classic marketing and monetary interest vs. reality. Believe what you want to believe though. Consciousness is flexible and consultative. Computer analysis is 1 dimensional and lacking responsiveness and creativity.

Go ask a surgeon how much he'd like a CT read by an AI. Seriously, you guys literally do not know what you don't know. It's a huge problem here that you can't even fathom, which is why you keep going on and on.

You have not made a convincing argument at all. All you can say is "this field is too complicated for you clods, so you are naive enough to believe the Radiology industry publications that say Radiology is being transformed by AI".

An AI can scan images thousands of times faster than a radiologist. That does not mean the radiologist has no other tasks to do (as has been explained to you multiple times before).

CAD in radiology has been around for more than 30 years.

Read up on concept of the exponential nature of technology. No one who can discuss futurism competently would present your sentence above as proof that technological progress was not occurring.

Again, Radiology publications themselves say AI is transforming radiology :

"If you think AI cannot replace radiologists, think again."

[Palamas]

Kartik,

Let me start off by saying that I am neither trying to be a thorn in your side, nor am I stating that computers will play no role as we move ahead. The way you talk about it is extreme though, in its application, and you have no evidence for it whatsoever, which is why your arguments are equally unconvincing. My arguments are nuanced. Let me explain a few, please:

First, if you can't sue anyone, in this system, it is a huge barrier. I'm sure other clinicians would love to blame a machine for the diagnosis or what they should do. What does that make it for them? Harder if a human isn't on the other side of the diagnosis to consult with. Telerad companies usually lose contracts currently or don't have the support of ER staff and other doctors on call (surgeons) for precisely the same reason. AI gives a 1 line dictation of dire diagnosis, correct or incorrect, and now it MUST be acted on, or considered, without discussion or flexibility. People don't like doing that, for 30 reasons I can't get into here. But you guys also think the robots will be doing the surgeries (LOL) so I digress.

I agree with the idea of exponential tech, but you can't just say it and then expect others to bow down. 30 years and it hasn't changed mammo. At all. Please explain why it wasn't exponential there. It wasn't even linear. It literally didn't change mammography at all, except for ... the same amount from when it came out til now, 30 years later.

Radiology publications, like the media, may or may not reflect reality. First radiologist were warned about outsourcing, then they were warned about "value", now they are warned about AI. Dr. Schier has no more predictive ability than others at the journal of ACR, or Dr. Siegel, who I've met:

https://www.carestream.com/blog/2016/11/01/debating-radiologists-replaced-by-computers/

Just so we aren't going back and forth, what exactly are you saying that you think I'm going against or contradicting?

Without radiology in modern America, hospitals basically shut down, by the way.

[Geoman]

Let's just ask the basic question - how long did it take for any technology to be adopted by 50% of the potential users?

Telephone - 45 years.
Computer 45 years.
Electricity 25 years.
Dishwasher 45 years.
Internet 25 years.
Microwave 20 years.

Transitions take time. Decades. Often the first 10-20 years sees small growth - the curve steepen quite a bit going forward.

Why does this happen? Inertia of the consumers - people stuck in the old way of doing things. Supply chains. Lack of enabling technologies (cell phones never really took off till batteries caught up). People turned off by early versions of the tech. Sometimes it takes a generation to literally die off so the new tech can be adopted by people not stuck in the past.

Your objections - well, you are completely misunderstanding what anyone has said.

Let's start with nuance here. When someone says "a robot" will do the surgery, they are not saying, without any human oversight, intervention, or control. You seem to be envision Rosie the robot with a scalpel. No one is saying that, no one has said that. I doubt anyone thinks that. It is straw man objection that you made up out of whole cloth.

I doubt an entirely autonomous robot will ever do surgery. At least for 40 years. I doubt anyone disagrees with that. What I expect will see is slowly increasing levels of computer enhancement and aid in surgery.

Blame the machine? My comment on no lawsuits was that you cannot sue for FREE services provided - I didn't specify whether those services were from a machine, a person, or a machine plus a person. I'm saying once the cost to do the examine has fallen, there MIGHT be a value in making it free, and changing the laws so that free consultation, which we ALREADY PROVIDE in numerous other fields could be done without accruing any liability. I'm not saying that can or will happen. I'm saying it might be a nice outgrowth of lowering the cost of the initial examine via partial automation.

Again, look around - seems most places are offering free exams with purchase of glasses. That is because they have, via technology, lowered the cost of the exam to a negligible expense. I see no barrier to doctors doing the same thing except medical liability laws. Change those might spur more innovation, and reduce costs for everyone.

By the way, the telemedicine situation we are discussing is already occurring in Alaska, where a clinician uses a nurse and a camera to examine people in remote villages.

https://www.healthcarelawtoday.com/2016/07/13/alaska-enacts-new-telemedicine-law-what-providers-should-know/

So we are not even discussing anything new, we are discussing something that is already happening, happening even more.

"Please explain why it wasn't exponential there. It wasn't even linear. It literally didn't change mammography at all, except for ... the same amount from when it came out till now, 30 years later." Simple - progress is exponential. That is obvious. The VCR was clunky, expensive, and almost unusable for the first 15 years - only 10% of consumers bought one during that time. And lives weren't literally at stake in recording TV shows. I expect medical devices, especially those involved in life or death diagnosis, will have long, slow, and low adoption curves. At least at first.

You know when rapid adoption happens in medicine? When someone sues the doctors because they refused to buy and apply the latest advance, and someone dies because of it. "Doctor X, is it true that if you had only used the latest diagnosis technology, my client might be alive today?"

The MRI was first invented and marketed in 1971. 12 years later there were 43 MRI scanners in the U.S. Two years after that there were 371. Now there are 13,000. The change is very slow at first, taking decades to even be realized. Then seemingly overnight, the technology is everywhere. The pattern has been repeated over and over again. So saying, Gee the tech was invented 30 years ago and it hasn't changed a thing....that is not actually all that surprising or unusual. In fact that is the typical. It means when the change happens, it will be even faster.

In my own industry, I was taught in school that fracking shale was pointless, and would never yield a thing. That was 1982. Flash forward to 2010 (28 years), and my fellow geologists were still calling fracking a flash in the pan, a financial Ponzi scheme, something that would never do what they promised. Then...the shale gale boooooom. now it is half the gas on the market. We were busy building gas import terminals for the coming gas shortage. Those are now export terminals.

Nothing changed, until everything changed.

[Joe Blow]

Great stuff Geoman,

The exponential growth of solar power is another case in point. If you look at the following wiki article there is a graph showing solar power growth from 2006 to present. It’s exponential. Before 2006, going back to the 1970’s, growth would look basically flat to the casual observer.

https://en.wikipedia.org/wiki/Growth_of_photovoltaics

The funny thing is, solar is still only 2% of world energy supply so it is still possible for a great many people to scoff at the idea that solar power will ever be significant. Yet it will only require another few doublings @ 30% growth rates (ie 5 - 15 years) to be highly significant.

The other thing about exponential growth is that for all the angst created by subsidies and other interferences in the solar market, the net effect is that the growth of solar has only been pulled forward by a few years at most.

[Kartik Gada]

Joe Blow,

You were reading my mind about Solar, which was about to be one of the upcoming ATOM AotMs.

It took 40+ years for PV to become 2% of the World's Electricity supply.....which means we are just 3-5 years from a major tipping point in solar. It is growing 30%/year, so 2% will be 4% and then 8% very soon. Plus, this is electricity that is going to be generated in very poor countries (sunnier countries are poorer) where there are still electricity shortages and unreliable grids.

[Kartik Gada]

A couple of Geoman's comments on this thread are just superb.

[Geoman]

If you really look at the curves, 5% market penetration seems to be the point where technologies start to take off. Usually it takes longer to get to 5% than it does to get to 50%.

I like to think that every technology has an ecosystem associated with it - a series of industries, sellers, suppliers, and other enablers that make it possible for a wide scale roll out. In some cases the old ecosystem has to actually die, or be cannibalized, before the new system can be born. And sometimes that takes a generation to happen.

Look at DVDs. They are a dying industry, yet how many people still have some DVDs and a DVD player in their living room? The death of Blockbuster means that ecosystem is failing, and Amazon prime and net flicks is the new ecosystem rising. But it is going to take some more time before things have fully transitioned. Maybe another 10 years? Eventually they will stop making players and disks, and more and more content will be on-line only. And the ecosystem will die, and the new system will finally reach full market penetration. or DVDs will live on, like the coelacanth, in a small corner of the technological universe.

The same transition is just starting with electric cars, but it is going to be decades before it is complete. The level of reordering, the petroleum fueled vehicle ecosystem is just too large. It will take decades to die. Which, given it took up >100 years to build it is not surprising.

Medicine is different from everything else, yet it is also exactly the same.

In many cases it comes down to sunk capital costs. When the sunk costs are high, change goes slowly. We have to unwind all that capital. And our sunk costs into the current system of medicine is immense - think of all those doctor's, with all those medical school loans that need to be paid off. Plus doctors expect that their sacrifice will be repaid in highly valued, well paid positions. For many elements of medicine that whole supporting ecosystem will have to change.

And thanks for the compliment Kartik! I find reading and commenting here helps clarify my own thoughts.

[Kartik Gada]

Geoman,

Part of why it was assumed that other countries could leapfrog the US is that there was not a big incumbent industry to unwind, where establish players try to obstruct new technologies. At the same time, a more primitive country may not have the ability to adopt the new technology either.

China was ideally suited to leapfrog, as it was still a shockingly underdeveloped country as recently as the 1990s, but had the ability to adopt any and all new technologies. China has accounted for a huge portion of the entire world's growth since that time. Unfortunately, the remaining countries that are still poor may not be capable of leapfrogging at all (or at least not quickly enough to qualify as 'leapfrogging').

I think Electric Cars can advance more quickly than that, since while it is true that the existing oil/gasoline network exists, the electrical outlet network also exists in homes, and is thus *more* convenient. Of course, >70% adoption is only by 2032, as discussed before.

Solar : The great thing here is that the highest upside is in places where the existing electrical grid is the worst (the sunniest countries are the poorest). We are at 2% now, but as PV grows 30% a year, we could reach 4% and even 8% before long, which is an immense tipping point.

Medicine : Doctors may protest, but they have no more power than any other profession being displaced. In fact, the regulatory rigging that inflated doctor compensation even beyond what is seen in other Western countries is what speeds up the disruptive force.

This is why ATOM-DUES is needed - the disruption of some jobs is cushioned by the exponentially rising DUES.

[Palamas]

Interesting responses, I agree in principle with some of these (that's why I'm here, I'm clearly not your foe in all regards) but I like to force you guys to temper your boldness. For example,

"The MRI was first invented and marketed in 1971. 12 years later there were 43 MRI scanners in the U.S. Two years after that there were 371. Now there are 13,000. The change is very slow at first, taking decades to even be realized. Then seemingly overnight, the technology is everywhere. The pattern has been repeated over and over again. So saying, Gee the tech was invented 30 years ago and it hasn't changed a thing....that is not actually all that surprising or unusual. In fact that is the typical. It means when the change happens, it will be even faster."

MRI (and even CT for that matter) is a case where the technology and improvement got better, yes, and that is WHY it was adopted to a great degree, and continually. What happens when the tech DOESN'T get better? You don't mention it as a failure, you just appeal to other things. That's why my CAD example for mammography is such a good example; it exposes your outcome biases. It shows that in particular areas, you appeal to this change, regardless of the subject, with a religious like devotion. I think this is example is one of them. Did you read Dr. Siegel's destruction of the opposing points in his article? He will retire in just a decade or so, so he doesn't have anything to worry about job wise or in terms of motivation for self preservation. Just saying. (400 pixel images? LOL)

The sunk capital costs is a good, important point. It also informs one how hard it is to get things through, especially if they aren't ready. Let's face it, the barriers to entry here are STEEP and the tech is not even close to anything relevant in the real world. I repeat NOT EVEN CLOSE. There are not even enough databases of images for training of computers, yet you get articles like the "Radiologists should worry" ... I'm laughing my ass off it's so outlandish. Half of the currently radiologists will be retired or dead before they even have a decent annotation or test set of images to TRY to train a machine to do anything near what a radiologist does. This is where scientism is strong in much of y'all's posts.

Electric Vehicles I could see being a way bigger blast off point for reasons Kartik has said, including social, practical, and cost points of view. Oil will continue to be used along side, but they might race against one another within the next 20 years, even much before the 50+ year supply of oil is even near running out ...

best to you guys

[Geoman]

When the tech doesn't get better? Well, you have cell phones, which for a decade were mostly crap. MRIs weren't very good the first 10 years. They certainly didn't sell very many. TV was invented at the turn of the century, and didn't become widespread for 30-40 years. I'm sure people thought those would never be a thing at some point. I actually had one of the first cell phones - came in a box like luggage. terrible reception.

Electric cars - you know we had electric cabs in Washignton DC in 1907. ttps://www.curbed.com/2017/9/22/16346892/electric-car-history-fritchle

So the tech there didn't get much better for 100 years.

You are asking us to predict when a technology that exists today will become good enough for widespread adoption and changing the profession. That is a vey tricky thing. Most people would fall back to just saying...soon. Computers are getting better all the time, as is imagining. When it crosses a threshold....? This seems like technology very susceptible to technological change.

Thing is it is always obvious in retrospect. We provide other examples because we know when the change occurred and and how, and we are trying to use those to predict when it might occur in your case. But it is imprecise, I admit. But saying it will never happen...well I used to say that about fracking, and I was right...for 25 years. A generation. Then I was suddenly very wrong.

I don't have a special devotion to accelerating change in technology - it is just an observation that it seems to be occurring. In fact it is so commonplace, we puzzle over instances where it is not occurring. We expect things to improve substantially in our lifetime. We don't realize that for most of human history very little changed. It used to be hundreds of years between useful inventions.

Kartik has a reasonably good track record - read some of the pages from 2008. he's made some dead on predictions. off by a few years in some cases, but the direction seems assured.

I don't believe in scientism, at all. In fact I am anti-scientism. Basically Scientism means that science should be used to determine values. heck, I'm not even enthusiastic about many of the changes coming - some I think might be detrimental.

And thanks for engaging in a conversation! I don't have even a smidgen of animosity toward you or your profession. I doubt Kartik does either. I certainly appreciate your perspective.

Be well.

[Kartik Gada]

I just attended an 'AI in Radiology' conference in Menlo Park, where quite a few MD Radiologists were excited about what AI was going to do to Radiology. The MDs weren't that young either. Some were in their mid-40s.

They agreed with me when I said that if MRI costs drop 10x, and precision rises 10x, then there will be so many MRIs that AI is needed to process them, and that many early-stage cancers will be caught (and percutaneously zapped) long before they are allowed to become more advanced.

The problem today is that a lot of cancer patients didn't even know they had a cancer until it was Stage III. Lower cost/more precise MRIs are the key to early detection. The much greater production of such images when the cost is this low means a lot of data for AI to train on.

[Andrew]

On the subject of tech deflation, there has been news lately that the US inflation rate has started declining again in recent months after a marked increase. This chart shows the recent downward trend:

https://tradingeconomics.com/united-states/inflation-cpi

I wonder if this is a consequence of the apparent overall stop in global QE that's also taken place recently, per the graph on page 3 of this document, showing the percentage change in major central bank assets going down to zero recently:

https://www.yardeni.com/pub/peacockfedecbassets.pdf

It's almost as if one trend follows the other, and there is almost a direct correlation.

[Geoman]

Well, a decline in fuel prices seems to be the cause of the inflation decline. Which goes to one of Kartik's older predictions - that oil prices have essentially topped out, and there is a barrier beyond which it becomes very hard for oil to sustain a price increase. Mostly due to the massive resources available from fracking, and the availability of new alternatives to fuel consumption.

That price wall for oil is likely to be declining as time goes by.

I could chart the same graph, but instead of comparing inflation to the fed assets, I could compare it to EV sales, which are increasing rapidly.

Unlike paste expansions, higher levels of growth are sustainable with less inflationary concerns. That is because technological deflation counteracts increased demands for goods and services. Inflation just drives technological deflation even faster.

It is a strange new world.

[Palamas]

Have you seen Raoul Pal's comments on the retirement crisis/demographics? He suggests that it's more about population booms and cycles than anything, re: inflation. It's funny because for totally different reasons, he's not worried about inflation at all, either.

The most important prediction has little to do with technology refining (this is, after all a net negative from a human point of view, imo) but rather when the next business cycle crash comes, and with it, a market crash that kills all these retirement accounts for good, with one swift stroke. It's not like millenial generations (Pal's point) will have anything to do with the market then, as it will be hard to find buyers at all to play this "just invest long term in the market for guaranteed gains" trope. Generally this makes sense, but we'll see if it shakes out that way. I think he's a bit off on ignoring how many other international investors will come to the US to find a way to escape their (more) broken bond and stock markets. My guess is middle of next decade. Until then, volatility, but still general boom. When Trump leaves after 2024, the future of the USA will be riddled with all manner of issues, even after all of his successes. Regardless, it was bound to come.

[Geoman]

I think retirement is part of it as well. But if the economy does poorer, then people will work longer, and with technological advancement, people might even live longer. So fewer young people, and more old people working into their 70s. “Difficult to see. Always in motion is the future..”

There are no truly poor countries. There is only good government and bad government, which explains every single outcome. Singapore should be poor, but is rich. Mexico should be rich, but is poor. It all comes down to economic freedom - there is an almost perfect correlation between the index of economic freedom and GDP per capita. And the government is 100% in control of the degree of economic freedom we enjoy. The U.S. improved its economic freedom last year by 0.6 points. This netted us around $200 billion in additional economic growth. Imagine what a couple of points might do.

This biggest hole in our pocket is government spending and fiscal health. But we have room to improve on property rights, financial freedom, and judicial strength. Just getting them to the same level as the UK or Australia would make a huge difference.

[Palamas]

Good post, Geoman, I agree.

Geo/Kartik/et al, what do you think is the reason why this narrative of growth at all costs (import foreigners who don't care to assimilate to your culture) exists, when you see the benefits of the Japanese model, regardless of if you think their bond market is a joke? Will they pay the price? But wouldn't you rather cycle and rebuild by maintaining your culture and country?

Kartik, why do you think the bond markets will be inept to fix the next crash? Is it because rates would have to be so high, they would paradoxically make a mockery of the broke status of the country = no one would take them. Basically, you'd have no real buyers and the confidence game is totally stalled?

Always good chatting with you all.

[Geoman]

Well, hmmmm.

I think the democrats have decided to get elected they don't need new ideas, they simply need a new electorate. So they decided to import one.

The republicans decided that to compete with China they need cheap labor, especially labor that falls outside of our cumbersome work rules and unions, so they decided to import one.

So both sides of the argument have strong reasons for supporting the existing and entirely insane system of immigration we have now. Both sides want there to be lots of illegal (not legal) immigrants.

Think about this - the obvious solution for Honduras or Guatemala would be to petition the U.S. to become states. There are only 9 million people in Honduras, 17 million in Guatemala - smaller than many existing states. They would exchange their largely corrupt government with one less corrupt, a stronger currency, access to the largest market in the world, and a strong military to keep the peace. Investment and the economy would soar in both countries. In a few short years there would be no need for anyone to immigrate anywhere - they'd be Hawaii of the south, full of rich retirees and vacationers.

Why is this never seriously considered by anyone? Well, because there is no real benefit to anyone important. The thugs that run these countries couldn't steal as easily. The cheap illegal labor would no longer be immigrating. And Dems would have no promise that the people of those countries would vote the "right" way. No, this wouldn't do at all.

Why doesn't Haiti petition to be a state? The Dominican republic? Jamaica? Why do pitiful, perennially corrupt, slow growth countries bother to continue as countries? It certainly isn't for the benefit of their people.

I'll let someone smarter then me debate the bond market. Well, and to observe, the less you need money, the more people line up to loan you money.

[Kartik Gada]

Andrew,

I wonder if this is a consequence of the apparent overall stop in global QE that's also taken place recently,

Yes. The fact that worldwide QE is falling, despite falling inflation, is a bad sign. This could lead to a recession (which I thought would be here by now, but could be soon nonetheless).

It's almost as if one trend follows the other, and there is almost a direct correlation.

There is. Inflation is falling again because Oil is declining for specific ATOM reasons (fracking and soon EVs). Hence, they should be doing more QE, not less.

[Geoman]

Amazon confirmed on Tuesday that it's launched a project to mine data from electronic medical records, as the company pushes deeper into the health-care market.

The company introduced Amazon Comprehend Medical, which "allows developers to process unstructured medical text and identify information such as patient diagnosis, treatments, dosages, symptoms and signs, and more," according to a blog post.

Sounds like machine learning and statistical analyses of diagnostic data from massive amounts of patients.

[Geoman]

Funny, I say it in November 2018, and by March 2020 it strat happening.

https://www.cnn.com/2020/03/03/business/walmart-health-care-clinics-amazon-cvs/index.html

"Health care looks like a big opportunity," Walmart CEO Doug McMillon told investors last month."We're not burdened by a legacy of health care service delivery," he added. "I don't own hospitals that I've spent billions on over-building, trying to figure out how to get a return on it."

[Kartik Gada]

Geoman,

Yes, you are right!

The ATOM always finds a way. The good news is that free market forces can pressure at least the basic GP visits. Two-thirds of the cars in any hospital parking lot are just basic GP/Ped visits.

[Geoman]

What about scanning and testing? The new trend - walk in MRI clinics. You get the MRI before seeing the doctor. It costs $400 versus $4,000. Not only do you save money, but you save a trip to the physician, him sending you for a scan, then second visit to consult over what they found.

https://expressmri.com/

Includes "Professional interpretation and a written report from a board-certified radiologist sent to you and/or your physician within 24 hours of your exam." My guess is they have a room full of very low cost radiologists that look at 20-30 scans per day from all over the country and write reports. Why not? Just like the eye doctor example I previously gave.

https://www.healthimaging.com/topics/healthcare-economics/fda-clear-worlds-first-portable-mri

Brings the MRI to you. And the equipment is 1/10th the cost.

[Kartik Gada]

Geoman,

It should happen. Many of the MRIs at Kaiser are done at a truck kept in the back of the building anyway. So why does it have to be through Kaiser's system at all?

That is even before any innovation in smaller size/portability. When it is more mobile, people getting MRIs proactively on their own will detect many cancers early (which is the whole point).